Split-Architecture OSPF Controller Intra-Area Path Cost Abstraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Split-architecture networks face sub-optimal routing due to the lack of accurate intra-area link cost information exchange between controllers, leading to inefficient path calculations and scalability issues, while traditional OSPF implementations either result in sub-optimal routes or excessive information exchange.
Innovation Solution
The implementation of a split-architecture OSPF protocol that abstracts intra-area path costs as direct links with cost values, allowing each controller to compute optimal paths independently, while maintaining backward compatibility with conventional routers and scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional OSPF implementations exchange complete intra-area link cost information between controllers, then routing accuracy is improved, but information exchange overhead and system complexity increase excessively
Solution Approach 1:
The patent extracts only the essential routing information needed for inter-area path computation - specifically, the shortest path costs between border switches - while leaving detailed intra-area topology information within each controller's local domain. This selective extraction achieves accurate inter-area routing without the overhead of exchanging complete topological data.
Solution Approach 2:
The patent segments the network into multiple areas with border switches identifying area boundaries. Each controller manages its own area independently, computing shortest paths locally and exchanging only boundary-related information. This segmentation enables scalable routing where each controller handles a manageable subset of the total network information.
2Device complexity
If split architecture networks use centralized control plane, then network management complexity is reduced, but routing optimization capability deteriorates due to limited information exchange
Solution Approach 1:
The patent divides the control plane into multiple distributed controllers, each managing a specific network area. This segmentation allows each controller to independently compute optimal paths within its area while exchanging summarized routing information with other controllers, thereby maintaining both management simplicity and routing optimization capability.
Solution Approach 2:
The patent changes the parameter being exchanged between controllers from complete topological details to aggregated shortest path costs between border switches. This parameter transformation enables controllers to make optimal routing decisions with minimal information exchange, balancing centralized management benefits with distributed routing intelligence.
3Loss of information
If all intra-area link cost data is provided to other controllers, then complete routing information is available, but communication overhead and processing time increase
Solution Approach 1:
The patent extracts only the critical routing parameters needed for inter-area path computation - specifically, the shortest path costs between pairs of border switches - and exchanges only this extracted information between controllers. This approach maintains complete routing information availability at the network level while minimizing communication overhead and acceleration convergence time.
Solution Approach 2:
The patent performs preliminary shortest path computation within each controller's local area before exchanging information with other controllers. This preliminary action pre-processes the topological data, transforming complex intra-area topology into simplified shortest path costs, thereby reducing the information exchange burden and accelerating overall convergence.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method is implemented in a network element that functions as one of a plurality of controllers for one of a plurality of areas of a split architecture network. The controller provides a control plane for the area of the split architecture network where the controller is remote from a plurality of switches providing a data plane for the area of split architecture network. The controller facilitates optimized routing across the plurality of areas of the split architecture network by providing limited intra-area link cost data to other controllers of other areas of the split architecture network and to traditional routers of a network including the split architecture network. The limited intra-area link cost data provides costs of each possible shortest path traversal of the area of the controller without providing all internal link cost data.